Polymer composites with the dielectric constant comparable to that of barium titanate ceramics

https://doi.org/10.1016/j.mseb.2007.01.035Get rights and content

Abstract

It has been found that CdO/polymer composites exhibit unusually high dielectric constant, much higher than that of BaTiO3/polymer composites, and comparable in magnitude to the dielectric constant of pure BaTiO3. While the dielectric constant of a BaTiO3/polymer composite at 1 kHz is below 20 at 20% volume fraction of BaTiO3, the dielectric constant of the corresponding 20% CdO/polymer composite is 2200. For both composites the relationship between logarithm of dielectric constant of the composite and volume fraction of the filler is linear within the range of useful filler contents where the composites are easily processable. Extrapolation of the data to 100% of the filler content yields the estimates of effective dielectric constants of pure fillers, which are of the order of 103 for BaTiO3 and 1015 for CdO. Thus, CdO behaves as a material with extremely high dielectric constant, when embedded into a polymer matrix. Preliminary data obtained for another composite based on tungsten dioxide (WO2) indicate that WO2 is another high-k filler.

Introduction

Rapid growth of electronic industry requires development of new materials with high dielectric constant (called high-k dielectrics) that would combine the high dielectric constant values intrinsic to ferroelectric ceramic materials with ease of processing characteristic for polymers. In particular, integration of resistors and capacitors into internal structure of printed wiring boards, or, directly into integrated circuits packaging requires materials compatible with the polymers used as supports of the electronic circuits [1], [2], [3], [4]. Organic polymers have relatively low dielectric constant, usually within the range of 2–10 [5]. In exceptional cases the dielectric constant of a pure polymer exceeds 10 (e.g., for poly(vinylidene fluoride) ɛ = 12,1 while for cyanoethylated O-(2,3-dihydroxypropyl)cellulose ɛ = 30 at 1 kHz), but always remains far below the dielectric constant of ferroelectric ceramics [6], [7]. In order to increase the dielectric constant of polymers, ceramic powders with high dielectric constant, such as barium titanate (BaTiO3), lead magnesium niobate–lead titanate (PMN–PT) or other, were added [8], [9], [10], [11], [12]. However, even at maximum filler loading the dielectric constant of ferroelectric ceramics/polymer composites rarely exceeded 100, because the ferroelectric ceramic/polymer composites with 0–3 type of connectivity (i.e., the composites where the filler particles are distributed at random in the polymer matrix [13]) follow closely an exponential relationship between their dielectric constant and the volume fraction of the filler [8], [9], [14], [15]. Logarithm of the dielectric constant of such composites (εcomposite) is linearly proportional to the volume fraction of the filler (φfiller) with the slope dependent of the dielectric properties of both components (Eq. (1)) [8].logεcomposite=φfillerlogεfillerεpolymer+logεpolymerAs the packing of a ceramic powder into a polymer is limited, usually the content of the filler cannot exceed the volume fraction of 0.40 or less, depending on the grain size distribution. The ceramic/polymer composites with higher than 40 vol% filler loading are hard to process into the form of thin films, required for high-density capacitors. Moreover, such composites exhibit extremely poor mechanical properties, which disqualifies them for application as embedded capacitor dielectrics. Hence, for example, to obtain easily processable polymer composites with the dielectric constant higher than 100 based on typical polymers with the dielectric constant of the order of 5, the dielectric constant of the filler must be higher than 9000. In order to match the magnitude of the dielectric constant of polycrystalline BaTiO3 ceramics, which is typically 1500–2000 [16], the effective dielectric constant of the filler has to greater than 106 at the filler content of 40% (by volume). To date, no such super-dielectrics have been known.

Recently, ultra high-k polymer composites have been developed based on metal or other highly conductive fillers, such as carbon black or carbon nanotubes [17], [18], [19], [20], [21], [22]. In these composites very high dielectric constant, of the order of 1000 or more, is observed only when the filler content is very close to its percolation threshold. It is believed that in the percolative systems the ultra high dielectric constant of the composites does not result from intrinsic dielectric constant value of the filler, but from giant interfacial polarization and effective increase of the electrode surface area due to formation of conductive particle clusters at the filler contents just below the percolation threshold [20], [23]. Far below the threshold the dielectric constant is small (i.e., less than 50), while above the threshold the capacitors made from these composites show extremely high dissipation factor or become simply shorted. In order to apply the percolative systems for manufacturing of capacitors, the conductive filler content and the filler distribution within the polymer matrix have to be precisely controlled, because less than 1% variation of the filler content exerts a dramatic effect on the dielectric constant magnitude and the magnitude of the dielectric losses [21].

In this paper, we report the first two ceramic fillers, which behave as materials with extremely high dielectric constant, when embedded into a polymer matrix. These fillers afford ceramic/polymer composite materials with the dielectric constant approaching and even exceeding the dielectric constant of pure barium titanate, and comparable in magnitude to the dielectric constant of the percolative systems.

Section snippets

Materials

Barium titanate (<2 μm, 99.9% of BaTiO3), cadmium oxide (∼1 μm, 99.5% of CdO), and tungsten(IV) oxide (<100 mesh) powders were purchased from Aldrich Chemical Co. and used as received. Trimethylolpropane triacrylate (TMPTA monomer) was obtained from Monomer Polymer & Dajac Labs. Benzoyl peroxide (initiator) was obtained from Aldrich.

Preparation of sample capacitors

The BaTiO3/TMPTA composites were obtained by photopolymerization of suspensions of BaTiO3 in TMPTA monomer, according to the procedures described previously [8]. The

Results and discussion

During study of various ceramic materials as potential fillers for polymer composites with enhanced dielectric constant, we have identified the first two ceramic materials that behave as materials with extremely high effective dielectric constant, when embedded into a polymer matrix. These materials are: cadmium oxide (CdO) and tungsten dioxide (WO2). When CdO or WO2 are incorporated into a polymer, the dielectric constant of the composites is very high, much higher than the dielectric constant

Conclusion

The first two ceramic materials with extremely high effective dielectric constant have been identified. When these materials are incorporated into a polymer matrix, inorganic/organic polymer composites are obtained, which have the dielectric constant comparable in magnitude to the dielectric constant of BaTiO3 ceramics. While the typical ferroelectric ceramic materials require high-temperature processing to make them in the form of thin wafers suitable for practical applications, the

Acknowledgements

R. Popielarz thanks the U.S. Department of Commerce and the Management of the National Institute of Standards and Technology for the financial support (a Guest Researcher compensation) received during the study of the composites in 1991–2001.

References (28)

  • R.E. Newnham et al.

    Mater. Res. Bull.

    (1978)
  • F.P. Koffyberg

    J. Solid State Chem.

    (1970)
  • P. Chahal et al.

    IEEE Trans. Comput. Pack. Manuf. Technol., Part B: Adv. Packaging

    (1998)
  • S.K. Bhattacharya et al.

    J. Mater. Sci.: Mater. Electron.

    (2000)
  • J. Prymark et al.

    Fundamentals of Microsystems Packaging

    (2001)
  • R.K. Ulrich et al.

    Integrated Passive Components Technology

    (2003)
  • B. Cherqaoui et al.

    Makromol. Chem., Rapid Commun.

    (1985)
  • T. Sato et al.

    Macromolecules

    (1991)
  • R. Popielarz et al.

    Macromolecules

    (2001)
  • C.K. Chiang et al.

    Ferroelectrics

    (2002)
  • S. Ogitani et al.

    Proc. Electron. Compon. Technol. Conf.

    (1998)
  • Y. Rao et al.

    J. Appl. Polym. Sci.

    (2002)
  • H. Windlass et al.

    IEEE Trans. Electron. Pack. Manuf.

    (2003)
  • Cited by (0)

    View full text